Volume 109: Where branding goes to die.
1. Where Branding Goes to Die.
tl:dr: Interbrand strangles three new GEs at birth.
Last year, perennial underperformer GE announced that in a Hail Mary pass of a turnaround strategy, it would split itself into three: An aerospace company, a healthcare company, and an energy business.
My, how the mighty have fallen. From the most valuable company on earth under Jack Welch, to bailed out by Warren Buffet under Jeff Immelt, to huge payouts to current CEO Larry Culp, GE and its shareholders have felt the pain. To put this into perspective, since 1999, the S&P500 has risen by 370%, while the value of GE has shrunk to less than 20% of its 1999 high ($82bn versus $451bn if you’re interested).
I’ll come back to why such a massive hit to corporate ego matters in a moment, but let’s shift gears to Interbrand for a second.
Interbrand won the golden lottery ticket when it was retained by GE to deliver the branding for all three new businesses. This is notable because typically, in a spin-out like this, each company to be spun out would retain its own agency partners, who’d then go off to work in isolation from each other. (Well done to Interbrand for that, BTW, it’s a bit of a new business coup)
We’re also incredibly fortunate that Interbrand’s NY CEO has been on the PR trail following the recent announcement of the work because it gives us an insider view into what they did and why they did it. I’m glad he did because we get to see why the work publicly announced so far is so utterly woeful.
Here’s a quote from said NY CEO Daniel Binns from this article [comments in square brackets are mine]. My apologies for it being a garbled mess. I’ve talked before about how Interbrand is incapable of communicating clearly ¯\_(ツ)_/¯. Anyhoo:
“Our overall approach to the branding decision was shaped by our proprietary Interbrand Thinking methodology…”
[I didn’t realize Interbrand needs a proprietary framework to think. This explains a lot]
“Specifically, we conducted a series of business case modeling simulations to calculate the net-present value (NPV) of future EBIT flows generated by each possible branding solution, in isolation for each company and in combination with the others (including the “base case” scenario of keeping the current GE-led industry-descriptive names for each company).”
[Sorry, this is nonsensical analytics theater. You can’t possibly break out branding factors like this and have anything like predictive accuracy. I’d love to see their margin for error, because it has to be huge. The read between the lines here should be “we padded our fees by a shit-ton to deliver the certainty of analytics theater because our clients were desperately afraid of screwing up. Again”]
“Our model started with baseline revenue inputs based on historical financials and future forecasts, and applied variable assumptions for brand-related factors with the potential to affect revenues and expenses going forward.”
[Ah, the needle in the proverbial haystack: “variable assumptions,” so this is yet another example of Interbrand doing fancy quantitative data analytics atop subjective assumptions pulled from its nether regions. This is basically its raison d’etre at this point]
“Monte Carlo simulations were then run to model a range of high-end and low-end assumptions for each variable factor, yielding a most probable financial outcome for the individual solutions and scenario combinations. This allowed us to compare the relative EBIT potential of each scenario, both in perpetuity and in the first five years post-separation.”
[Oh, Interbrand, you’re blinding us all with your intelligence. Put it away already. Also, note the term “EBIT potential of each scenario…in perpetuity.” Let’s be clear, such a calculation is literally impossible to undertake with any degree of accuracy and just makes them look silly for claiming they can]
Before I move on, I want to be serious for a second. Brand valuation, which Interbrand has done for years, is the underlying basis for everything written above. Now, a backward-looking and subjective valuation delivered using historical data to a fairly rough and ready degree of accuracy is one thing, but using that same method to make predictive recommendations is quite another. And quite dangerous if we consider the sheer scale of the error factor that must be inherent in these calculations. To put this error in perspective, both Interbrand and BrandZ value the VISA brand in their annual league tables, but Interbrand believes it to be worth 15 times less than BrandZ does–$15bn versus $191bn. Put simply, if two different yet “gold standard” methodologies using the same historical data can have a delta this large, then we cannot possibly take predictive, future-facing calculations based on unknown data using the same methodology seriously. And by attempting to do so, I’d argue that a business like Interbrand, far from de-risking the process, is actually being wildly irresponsible.
OK, so enough about valuation nonsense; what about the work? Well, Interbrand isn’t nicknamed Interbland for nothing. Aside from a predictable degree of mediocrity, three things are notable from what’s been shared:
All three businesses will be branded GE, and all three will have the GE Monogram attached. Why? Because the GE name and mark have equity and compared to a completely new name, it has 100% more. This is obvious and an analysis that could’ve been done in seconds, for free.
They color-coded the three brands in “evergreen,” “compassionate purple,” and “atmosphere blue.” Aside from those names being awful, this isn’t particularly smart. Color coding rarely, if ever, works because it assumes your customers are paying way more attention than they are. The only saving grace here is that each corporation will operate in different markets with different customers, so confusion is likely not that big of a deal, at least not initially.
They did something horrendous to the typeface. Three times. This is a shame because, alongside the name and the Monogram, the GE typeface is one of the few distinctive assets the GE brand has. At least one of the new businesses should’ve kept it. Sigh.
In the article quoted above, Mr. Binns says the whole process was equal parts analysis and creativity, but that appears to be false. Instead, it looks like 20% analysis, 80% analytics theater, and no creativity anywhere to be seen, asphyxiated as it was under the sheer weight of the other two. What a sad way to go. RIP branding. Well done, Interbrand.
Aside from Interbrand being where branding goes to die, I take two things from this. First, Interbrand is clearly so proud of the process–2,200 industry practitioners and 3,000 employees, EBIT Modeling, Monte Carlo Simulations, and interdependent and independent teams–that they appear unaware of the utter mediocrity of the outcome this process led to. In other words, this looks like a lot of time and money spent defensively analyzing the task to death because the goal was not screwing up, rather than working backward from the desired outcome of what it would take for each of these new brands to win in their respective markets. Second, in delivering such a defensive “cover your ass” approach, Interbrand is simply serving the client what it wants. The clients most likely to clamor for “data-driven” solutions like this and thus the most likely to hire partners like Interbrand are normally underperforming for some reason or another, and branding defensiveness and business underperformance are often connected. (Back to my point about GE being a company whose corporate ego and fundamental sense of self have taken a battering in recent years. This creates a bunker mentality where the fear of screwing up massively outweighs any desire to seize the day)
Confident corporations that outperform have a clear vision and strategy and don’t require the safety blanket of such patently flawed analyses. They don’t waste their time second-guessing every possible permutation and seeking to quantitatively analyze those things that are exceptionally hard, if not impossible, to measure, let alone predict. Nope. They focus their energy on what matters, where it matters, and how to take the business into the future confidently.
Now, I’m going to leave you with something that might seem a bit left-field, but hopefully, you’ll bear with me. First, look at the above, and then read what Roger Martin says in his latest post on the difference between planning and strategy. I think you’ll see the overlap and how what Interbrand did with GE is rather more technocratic than it is strategic. I’d argue that this might not be the most helpful thing for three new businesses about to embark on a whole new choose-your-own-adventure.
2. Lies, Damned Lies, and Marketing.
tl;dr: How much does it matter if a dataset is crappy?
Prof. Byron Sharp, head honcho of that merry band of Australian Marketing Scientists over at the Ehrenberg Bass Institute, popped up last week at a marketing conference to make some pronouncements.
If you’ve paid attention to his previous work over the past 15 or so years, none of what he said will have come as a surprise - mental and physical availability, check. Reach, check. Distinctive assets, check. (And if you haven’t been paying attention, you should, as “How Brands Grow” remains the single most important book in our field, even if I do take issue with parts and remain frustrated that because Prof Sharp is an academic, there’s a distinct lack of practical how-to in a book that purports to be).
Anyway, in what appeared to be a calculated takedown, he stated that the Binet & Field 60:40 “rule” for a media mix, where you put 60% of your budget into brand spend and 40% into activation, is nonsensical, primarily because of two things:
The dataset, being built around campaigns that win awards, is crappy.
It’s unclear what the difference is between brand and activation activities, so what are we measuring?
Now, before we move on, let me state that at an evidentiary level, I agree with the above two points and have struggled with both myself. However, from a practical standpoint, I disagree with his out-of-hand dismissal.
Here’s why.
First, there’s no such thing as perfect data in marketing since the systems we’re dealing with are complex and adaptive. In fact, the issue I take with academics like Prof Sharp is inherent in the idea that we can perfectly isolate and quantify specifics amid such complexity–reality suggests that it’s not so easy. Worse, all too often, attempts to isolate and quantify qualitative data result in what Tim O’Reilly dubs the “clothesline paradox,” which is our tendency to ignore and label unimportant those things that are hard, or impossible, to measure quantitatively–like how much energy is used drying clothes on a washing line compared to the tumble drier. This begs the question of whether the awards nomination dataset is any better or worse than any other and whether, ultimately, it matters all that much? My take is that roughly right is generally good enough in most instances and that it’s wrong to dismiss roughly right simply because it isn’t perfectly correct.
Second, as a practitioner, there are times when you need to separate yourself from academic rigor in the interest of what’s practically useful.
I’ve had many conversations with clients around 60:40, and it’s been useful every time. When I discuss it, I don’t use it as a “rule” but instead refer to it as a rule of thumb that’s broadly applicable as long as we don’t take it too literally (every client situation is different, I’m not a media expert, you should talk to someone who is, etc.) It’s a general statement that brand building matters. Where this conversation is most pertinent - and has the biggest impact - is when you’re talking to clients where all, or almost all, of their current budget goes toward performance and programmatic activities; in other words, heavily biased toward activation.
With these clients, blending the 60:40 conversation with a discussion of top-of-funnel activities, reach, distinctiveness, memory structures, mental and physical availability, salience, awareness, positioning, etc. all come together to help demonstrate the context in which brands help drive business success.
Sadly, the trade press picked up primarily on the 60:40 takedown, which meant that somewhat hidden was a much more important pronouncement that you’d be forgiven for missing entirely.
And this is that “volume creative,” AKA spam, really is a waste of your budget. That’s right, while Prof Sharp was busy dunking on folks he knows are broadly in line with his own thinking, he was also pointing out something vastly more important - that snake oil sellers like Gary Vee truly are full of shit.
3. “Call it Something New. Call it Innovation.”
I used to joke back in the day that nobody in our business did any actual innovation; we just changed the names of things and called it innovation.
This is one of the reasons that if you get a bunch of agency people and marketers in a room, you need to start the conversation with an agreed-upon glossary of terms because, without it, they’ll all be talking at cross-purposes. It’s so bad that, at times, I’ve found myself walking clients through proposals and sharing all of the other terms that might be used to describe the same deliverable or process step so they’re better able to understand what they’re looking at.
Now, at face value, this is obviously kind of silly, but recently Econometrician Dr. Grace Kite made an important point that this lack of precision can also be harmful.
Here, she points out issues with terms such as ROI, which remains squirrelly when applied to marketing (In any business school, when you mention ROI, the correct next question is “over what time period?” A question marketers would do well to ask more often), and “cost per acquisition,” which she describes like this:
“At best it’s a grave misunderstanding. More likely it’s a flat out lie, told because it benefits the platforms that report it.”
And then goes on to describe issues with incrementality.
While we’re at it, it’s a shame she didn’t hit on things like “performance marketing,” which is called that because it’s paid for based on a measure of performance, like a click, rather than because it performs any better or worse than any other marketing activity. Or “growth hacking,” which I once had a client define as “marketing, just without any ethics,” which I chuckled at.
Anyway, I agree with Dr. Kite. We have a terminology problem; often, it leads to conversations that run at cross-purposes, but more concerningly, the everyday use of inaccurate terminology can sometimes be flat-out misleading.